By now, the comparisons between the Iowa Caucuses and the Electoral College are becoming apparent. Bernie Sanders has been leading in tallies of the votes, both the first round and the realignment. But Pete Buttigieg is winning more state-delegate equivalents, the Iowa Democratic Party's preferred measure, and the measure more closely connected to the purpose of the caucus, which is to select delegates to the county caucuses next month.
The reason this can happen is the same reason the Electoral College can give the win to a candidate who doesn't get the most votes. A surplus of supporters in one state are wasted, and those who fail to reach a threshold (majority for the Electoral College, viability for the caucus) don't count either. It matters not just who you vote for, but where.
For the Electoral College, one can make the case that our federal democracy is a union of states, so representation should flow through the states. But it's less clear that we should think of Iowa as a union of precincts.
But the problem is worse than for the overall winner.
Much of the drama out of the caucuses involves the failure of some candidates to be viable in some precincts. If they fail to see the threshold, they're out, and their supporters realign to another candidate.
This sort of ranked-choice voting seems appealing. If your preferred candidate can't get much support, you get to express your second choice. But realignment isn't just for marginal candidates. It's for candidates who miss the threshold in any particular caucus. So small candidates with dispersed support are disadvantaged over small candidates with geographically concentrated support.
I spent caucus night floating between three caucuses in Iowa City. (for more observations from those caucuses, see here). Each caucus had around 500 participants, meaning more than 70 caucusgoers were needed for viability.
In two of those precincts, Amy Klobuchar was viable, and came away with delegates. In the other, she had 50 supporters on the first alignment, 29 shy of viability. After bargaining with 52 Biden backers and 57 Buttigieg supporters, some (but not all) of the Klobuchar supporters switched to Buttigieg, and he reached viability.
Had those Klobuchar supporters been in another precinct, or if some from the neighboring precinct had been in this one, she'd have one another delegate.
Had those Klobuchar supporters been in another precinct, or if some from the neighboring precinct had been in this one, she'd have won one more delegate. Meanwhile, had all of Biden's supporters across all three precincts been in the same precinct, they would have dominated.
One solution would be lower thresholds, which would probably require combining precincts, so each precinct had more delegates to award. That would undermine one of the selling-points of he caucus, the very personal nature of the small-group discussion. But these caucuses were already at 500 people, so discussion was limited.
Political science emphasizes the importance of district magnitude in explaining how proportional a system is. We often focus on how the ballots are counted — ranked-choice-voting, proportional allocation, and so on — but the real variable is the size of the district. If the Democrats want to continue to value proportionality, they should consider focusing on it.